 Alexander Korolev
|
17a4f0c39a
feat(i18n): localize all remaining module output strings
|
4 months ago |
|
Alexander Korolev
|
54de1eb75d
fix: restore VLLM_LOGGING_COLOR=1 alongside TERM=xterm-256color
|
4 months ago |
|
Alexander Korolev
|
679c256a60
refactor: resolved_hf_cache_dir returns PathBuf instead of String
|
4 months ago |
|
Alexander Korolev
|
cd8d9d66d5
refactor: replace wildcard imports with explicit imports (docker/mod.rs)
|
4 months ago |
|
Alexander Korolev
|
68132ce91a
feat: set TERM=xterm-256color on head node for colored logs
|
4 months ago |
|
Alexander Korolev
|
227eb18915
feat: auto-set VLLM_LOGGING_COLOR=1 on head node (override via recipe [env])
|
4 months ago |
|
Alexander Korolev
|
7b01e19ec2
fix: HF_HOME -> /root/.cache/huggingface (fixes stale cache)
|
4 months ago |
|
Alexander Korolev
|
9241bdc682
fix: don't pass --distributed-executor-backend mp (auto-detected)
|
4 months ago |
|
Alexander Korolev
|
6f2c33a82f
fix: only add --shm-size when recipe explicitly sets it
|
4 months ago |
|
Alexander Korolev
|
d9c3f845c2
fix(deepseek): align parameters with working system
|
4 months ago |
|
Alexander Korolev
|
2af1104e7f
feat: launcher auto-adds --host/--port; recipes no longer hardcode
|
4 months ago |
|
Alexander Korolev
|
7ba894be67
feat: per-recipe shm_size + DeepSeek V4 Flash recipe
|
4 months ago |
|
Alexander Korolev
|
9c0e7680a4
feat: set safe NCCL defaults (IGNORE_CPU_AFFINITY, NET, CROSS_NIC, DEBUG)
|
4 months ago |
|
Alexander Korolev
|
1b02f53b07
feat: set GLOO_SOCKET_IFNAME and TP_SOCKET_IFNAME for vLLM
|
4 months ago |
|
Alexander Korolev
|
bfe94fa39e
feat: pre-serve hooks + exec subcommand for cluster mass-execution
|
4 months ago |
|
Alexander Korolev
|
5097d55c77
fix: redirect stderr in `cargo run logs --follow` docker pipeline
|
4 months ago |
|
Alexander Korolev
|
d92da29eb5
fix: shell-quote entrypoint script so 'sh -c SCRIPT' survives bash word-split
|
4 months ago |
|
Alexander Korolev
|
75860da26d
feat: test-sleep recipe streams logs; add --timestamps to docker logs
|
4 months ago |
|
Alexander Korolev
|
2b0be29648
feat: cluster defaults, configurable init_port, --headless for workers
|
4 months ago |
|
Alexander Korolev
|
afd2fc35bf
feat: switch SGLang to --dist-init-addr, add memlock/stack ulimits
|
4 months ago |
|
Alexander Korolev
|
8194da83d4
fix: stop/logs use CLI recipe name, health uses correct label filter
|
4 months ago |
|
Alexander Korolev
|
4d3c297c50
feat: add --rm to containers by default, --keep flag to persist
|
4 months ago |
|
Alexander Korolev
|
91e273a649
feat: health command shows all containers, marks ours (sm121)
|
4 months ago |
|
Alexander Korolev
|
a726b83bcf
feat: mount HF cache directory in Docker containers
|
4 months ago |
|
Alexander Korolev
|
38290ec6f5
refactor: split docker.rs and tmux/mod.rs under 500 lines each
|
4 months ago |
